21 results about "Common noise" patented technology

The invention discloses a drilling multi-source sensing signal denoising method based on dynamic noise decoupling and a self-adaptive mode, and belongs to the technical field of underground processing. The method comprises the following steps of: separating common noise of a noise energy distribution matrix of a sensor and separating specific noise; carrying out adaptive noise set empirical mode decomposition on the separated signals, carrying out variational mode decomposition on residual signals in the signals, and screening effective modes through a kurtosis-entropy joint criterion to obtain signals with effective characteristic components reserved; high-frequency fluctuation of the signal is punished through total variation regularization, an improved alternating direction multiplier method algorithm is used for solving, short-time Fourier transform is carried out on the solved signal, low-frequency and high-frequency features are extracted through a multi-scale convolutional network, and a final denoised signal is output through gating weight fusion. According to the method, the signal denoising precision in a complex noise environment is remarkably improved, the processing time is shortened, and the resource consumption is reduced.

The invention discloses a combined preprocessing and fusion decoding method based on electroencephalogram signals and functional near-infrared signals, which comprises the following steps of: synchronously acquiring the electroencephalogram signals and the functional near-infrared signals, taking a task trigger event as a unified time reference, and carrying out basic preprocessing of time alignment and modal self-adaption on two modal signals to obtain a functional near-infrared signal; a pre-trained cross-modal noise joint modeling module is utilized to perform joint modeling on cross-modal joint noise features caused by a common noise source, collaborative denoising processing is performed on multi-modal signals based on the cross-modal joint noise features, multi-modal denoising representation is obtained, a cross-modal feature interactive modeling mode based on an attention mechanism is obtained, and the multi-modal noise is obtained. And dynamically modeling the correlation between different modal features, realizing deep fusion of multi-modal complementary information, carrying out decoding processing based on the fused features, and outputting a corresponding brain-computer interface control instruction or task identification result. According to the method, the decoding accuracy and robustness of the brain-computer interface system in a complex task scene are enhanced.

The invention relates to a high and steep slope real-time monitoring method and system based on multi-source data fusion and cooperative enhancement. The method comprises the following steps: scanning a slope provided with a corner reflector array by using a synthetic aperture radar (SAR); performing real-time differential correction on data acquired by the radar by taking the corner reflector in the stable region as a reference so as to eliminate common noise; meanwhile, a three-dimensional laser scanner is used for obtaining point cloud data of the slope; coordinate registration and fusion of radar data and point cloud data are realized through the corner reflector array, so that a three-dimensional deformation field of the slope is generated. The method provided by the invention has the advantages of high monitoring precision, high signal-to-noise ratio and good data reliability.

The invention discloses a synthetic aperture radiometer darkroom phase calibration method and device. The method comprises the following steps: installing a synthetic aperture radiometer and a point target emission source in a microwave anechoic chamber; in the first-stage test, unit antenna output ports are connected with equal-length test cables and connected to a microwave test instrument, a synthetic aperture radiometer is set to be in an antenna observation mode, and a point target observation unit antenna phase difference test is carried out; in the second-stage test, the output end of the unit antenna is connected with each receiving channel of the synthetic aperture radiometer, the synthetic aperture radiometer is set to be in an antenna observation mode, and a point target observation full-system phase difference test is carried out; in the third stage of testing, the synthetic aperture radiometer is set to be in a common noise injection calibration mode, and common noise injection calibration is carried out; and calculating and counteracting a near field effect phase difference, a unit antenna phase difference and a receiver channel phase difference to obtain an accurate noise distribution network phase difference parameter. Darkroom phase calibration is achieved under the near-field condition, the result is accurate, and the method is simple, convenient and practical.

The invention discloses a full-band voice noise reduction method, device and equipment based on a deep neural network. The method comprises the following steps: generating a noisy voice data set according to a preset noise audio file and a preset pure voice file; constructing a full-band noise reduction model, and training the full-band noise reduction model by using the noisy voice data set to obtain an optimal full-band noise reduction model; and using the optimal full-band noise reduction model to perform noise reduction processing on a collected to-be-processed voice signal. According to the technical scheme provided by the invention, the full-band noise reduction model based on the deep neural network is constructed and trained for noise reduction processing, the noise reduction processing of the full-band voice signal is realized, most of common noise in life can be inhibited, the inhibition effect on non-stationary noise is obvious, the overall residual noise is lower, and the noise reduction effect is better. And the quality, definition and the like of the voice after noise reduction are remarkably improved.

The invention belongs to the technical field of geological exploration, and particularly discloses a symmetrical excitation anti-explosion geological exploration system and method, and the method comprises the steps: receiving array multi-channel signals at two sides, and carrying out the time alignment correction of all channel signals, so as to eliminate the time delay; a multi-channel cross-correlation matrix between the left array and the right array is constructed based on the aligned signals, and matrix elements are obtained through integral cross-correlation operation of array channel signals on the two sides; performing low-rank decomposition on the cross-correlation matrix, and extracting the first k main components as public noise modes; carrying out inversion calculation on the contribution of each noise mode in each acquisition channel, and reconstructing a channel noise component; and subtracting the reconstructed noise component from the original signal, and outputting a purified geological reflection signal. The method does not need special hardware, can accurately suppress the homologous low-frequency noise of the tunnel, improves the signal-to-noise ratio of the signal and the geological imaging resolution, and is suitable for geological detection of closed spaces such as the tunnel.

The application discloses a joint preprocessing and fusion decoding method based on electroencephalogram and functional near-infrared signals, comprising the following steps: synchronously collecting electroencephalogram signals and functional near-infrared signals, and taking a task trigger event as a unified time reference to perform time alignment and mode adaptive basic preprocessing on the two kinds of mode signals; using a pre-trained cross-modal noise joint modeling module to jointly model cross-modal joint noise features caused by a common noise source, and performing cooperative denoising processing on the multi-modal signals based on the cross-modal joint noise features to obtain multi-modal denoised representations; using a cross-modal feature interaction modeling mode based on an attention mechanism to dynamically model the correlation between different mode features, realizing deep fusion of multi-modal complementary information, performing decoding processing based on the fused features, and outputting corresponding brain-computer interface control instructions or task recognition results. The application enhances the decoding accuracy and robustness of the brain-computer interface system in a complex task scenario.

This invention provides a single-phase three-wire grid-connected inverter circuit that can suppress common noise more effectively than conventional circuits. [Solution] A single-phase three-wire inverter circuit comprising: a voltage divider circuit comprising two capacitors, with the neutral AC side connection terminal connected between the two capacitors; a first leg comprising two switching elements, connected to the first AC side connection terminal via a first inductor between the two switching elements; a second leg comprising two switching elements, connected to the second AC side connection terminal via a second inductor between the two switching elements; and a third leg comprising two sets of two switching elements, each with a diode facing in opposite directions, connected to the neutral AC side terminal via a third inductor between the two sets of switching elements, passing through the midpoint of the voltage divider path.

The invention relates to an atom interferometer with a laser light source through which first laser light can be emitted at a fundamental frequency, and with at least one light emission device through which second laser light can be emitted at a frequency that deviates from the fundamental frequency by a difference frequency, and with an atom trap in which one or more atom ensembles can be captured and irradiated with the first and the second laser light.

The utility model relates to a noise reduction type dry-type transformer which comprises a shell and a body arranged in the shell, a damping device is arranged at the bottom of the body, the body is arranged in an installation frame, the installation frame comprises transverse rods and vertical rods, the vertical rods are arranged at the two ends of the transverse rods, the two transverse rods are arranged in parallel, side plates are arranged between the adjacent transverse rods, the vertical rods are provided with connecting pieces, and the connecting pieces are connected with the shell. An adapting plate is arranged on the inner wall of the shell, the connecting piece is connected with the adapting plate through a locking assembly, the damping device is arranged at the bottom of the mounting frame and comprises an outer shell, a supporting part, a first damping part and a second damping part, the first damping part is arranged above the second damping part, and the first damping part and the second damping part are arranged in the outer shell; the supporting part is arranged above the first damping part and arranged on the upper surface of the shell in a protruding mode. According to the noise reduction type dry-type transformer, noise can be transmitted through the mounting frame and the shell, and noise generated by vibration and resonance noise are reduced.

The utility model relates to the field of workshop noise reduction devices, in particular to a zipper production workshop noise reduction device capable of being flexibly adjusted. The sound insulation cover further comprises self-locking pull rods, a U-shaped connecting frame, folding plates, connecting rods, extending covers and pulleys, a connecting groove is formed in the front side of the sound insulation cover, the four self-locking pull rods are symmetrically installed in the connecting groove, the other ends of the self-locking pull rods are connected with the U-shaped connecting frame, and the folding plates are symmetrically installed on the front side of the U-shaped connecting frame. The upper end of the front side of the U-shaped connecting frame is symmetrically connected with two connecting rods, and the other ends of the folding plate and the connecting rods are connected with an extension cover; according to the noise reduction device for the workshop, the extension cover is arranged, so that the coverage range of the acoustic hood can be flexibly adjusted, and the problems that a common noise reduction device for the workshop is not provided with an adjusting mechanism, the size of the acoustic hood cannot be flexibly adjusted, the size of the acoustic hood is easily not matched with that of a noise source, normal use of the acoustic hood is affected, and noise is easily generated are solved. And therefore, the practicability of the acoustic shield is reduced.

The object is to provide a measuring device and a measuring method and a measuring program thereof that can effectively reduce the influence of noise. The measuring device (1) comprises: N (N≥2) sensors (S0, S1, S2); a selection unit (4) that selects a given combination of sensors for each measurement and outputs a detection value to each of M (M<N) sensor terminals based on the measurement value of the selected sensor; a detection unit (5) that obtains the detection value output to each of the M sensor terminals for each measurement; and a correction unit (8) that, after performing L measurements, assumes that each of the M×L detection values ​​obtained by the detection unit (5) contains a time-dependent noise component and corrects each of the M×L detection values ​​so that the time-dependent noise component of each detection value becomes a common noise component, which is a value obtained by averaging the L time-dependent noise components in the time axis direction.

This invention provides a single-phase three-wire grid-connected inverter circuit that can suppress common noise more effectively than conventional circuits. [Solution] A single-phase three-wire inverter circuit comprising: a first leg corresponding to the first phase AC side terminal, comprising two switching elements and connected to a first inductor midway between the two switching elements; a second leg corresponding to the second phase AC side terminal, comprising two switching elements and connected to a second inductor midway between the two switching elements; and a third leg corresponding to the AC side terminal related to the neutral wire, comprising two sets of two switching elements each equipped with a diode facing in opposite directions and connected to the neutral wire AC side terminal midway between the two sets of switching elements.

This invention relates to the field of wind turbine technology and specifically discloses a noise reduction device for a wind turbine. It includes a fixed circular plate, a curved baffle fixedly connected to one side of the fixed circular plate, a circular sleeve plate fixedly connected to the side of the curved baffle away from the fixed circular plate, and a concave sound-absorbing cotton fixedly connected to the side of the fixed circular plate near the curved baffle. This noise reduction device for a wind turbine, by setting multiple curved baffles around the outside of the generator assembly to block and reduce noise, prevents the common noise-reducing shell from only sequentially blocking and intercepting noise waves, resulting in poor noise reduction and still causing a certain degree of noise. By setting the curved baffles with a zigzag shape around the outside of the generator assembly to block noise while facilitating air circulation, it prevents the sound-absorbing shell from completely enclosing the wind turbine in a sealed state, which would lead to low heat dissipation efficiency and be difficult to manage during long-term operation.

The invention relates to an atom interferometer with a laser light source through which first laser light can be emitted at a fundamental frequency, and with at least one light emission device through which second laser light can be emitted at a frequency that deviates from the fundamental frequency by a difference frequency, and with an atom trap in which one or more atom ensembles can be captured and irradiated with the first and the second laser light.

The utility model relates to the technical field of oxygenerator devices, in particular to an efficient noise-reduction oxygenerator which comprises an outer shell, a turning cover is assembled at the upper end of the outer shell, an air inlet is formed in the surface of one side of the outer shell, a base plate is assembled at the bottom of the outer shell, an air filter box is assembled on the inner side of the air inlet, and a sound absorption plate is installed at the rear end of the air filter box. The upper end and the lower end of the sound absorption plate are provided with the sound absorption side plates, the rear end of the air filter box is connected with the compressor, the rear side of the compressor is provided with the input pipeline, and the rear end of the input pipeline is provided with the molecular sieve adsorption tower. In addition, a closed environment made of sound absorption materials is assembled at each structure in the device, and a sound absorption layer arranged on the inner surface is matched, so that the interior generated can be greatly reduced, the generation of noise is effectively prevented, the device is more useful than a common noise reduction mode, and the interior of the device can be protected.

This invention discloses a method and system for synergistically improving the key performance of dynamic measurement of fiber optic current transformers, including: constructing a first mathematical model of forward channel gain and step response rise time; constructing a second mathematical model of forward channel gain and overshoot; constructing a third mathematical model of common noise and signal-to-noise ratio of the fiber optic current transformer system; based on the first and second mathematical models, with forward channel gain as the independent variable and minimizing overshoot and step response as the objectives, performing dual-objective optimization to determine the optimal forward channel gain, thereby minimizing the dynamic measurement performance (overshoot and step time) of the system; using the forward channel gain value obtained from the optimization as a constraint, improving the third mathematical model with the objective of maximizing it; and guiding the design of fiber optic current transformers based on the optimization results, which can effectively ensure that the fiber optic current transformer has good dynamic measurement performance while having a high signal-to-noise ratio, thus improving the measurement performance of the fiber optic current transformer.

The invention provides a double-spectrum iris image intelligent recognition method and system based on multi-stage characteristics, and relates to the technical field of iris image recognized.The method comprises the steps that firstly, a clear sclera boundary provided by a visible light image is used for guiding precise segmentation of a near-infrared image, and high-precision initial characteristic positioning is completed; according to the method, standard local binary features are extracted, a weight map capable of quantifying the spectral stability of each local area is generated by analyzing the difference of co-occurrence matrixes in the double-spectrum gradient direction, then collaborative filtering is carried out on the double-spectrum features in a frequency domain, non-common noise is removed, and the spectral stability of each local area is quantified. The method comprises the following steps: firstly, carrying out self-adaptive weighted fusion in a spatial domain by using a weight map to generate a highly robust composite feature, and finally, carrying out weighted Hamming distance calculation by multiplexing a spectrum stability weight map in an identification matching stage, so that the matching process focuses on a reliable feature region with high spectrum consistency. And the accuracy of the iris recognition system in the complex imaging environment is improved.

The utility model aims to provide an acoustic metamaterial sound absorption and sound insulation barrier aiming at the problems of the existing sound insulation barrier in a construction site. The acoustic metamaterial sound absorption and sound insulation barrier is composed of one or more sound absorption and sound insulation barrier units. Each sound absorption barrier unit sequentially comprises a front porous sound transmission surface, a waterproof sound transmission layer, one or more metamaterial sound absorbers and a rear sound insulation surface; or each sound absorption barrier unit sequentially comprises a front porous sound transmission surface, a front waterproof sound transmission layer, one or more bidirectional metamaterial sound absorbers, a rear waterproof sound transmission layer and a rear porous sound transmission surface. The working frequency spectrum of the acoustic metamaterial sound absorber of the device is designed for the common noise of the construction site, so that the overall noise is efficiently reduced; a traditional fiber porous sound absorption material is not used in the device, the problems of water accumulation, mildewing, layering pulverization and the like do not exist in the internal structure of the device, and the outdoor service life is long.

The invention discloses a cross-chain interaction method and device based on a block chain, equipment and a storage medium, and relates to the technical field of block chains, and the method comprises the steps: obtaining a first public key generated by a key generation center, acquiring a first encryption parameter and a second encryption parameter issued by the first block chain and the second block chain after encrypting the first negotiation parameter value and the second negotiation parameter value based on the first public key and the common noise value; judging whether a plaintext difference determined based on the first encryption parameter and the second encryption parameter is consistent with a preset plaintext difference value, and if yes, calling and obtaining a polynomial commitment, witness information and zero-knowledge proof determined by the first block chain based on the first negotiation parameter value; and verifying the polynomial commitment, the witness information and the zero knowledge proof, if the verification is passed, verifying a data exchange proof generated after data exchange between the first block chain and the second block chain, and if the verification is passed, completing a cross-chain operation between the block chains. Therefore, the data transmission security of the block chain can be improved.

The invention discloses a multi-type weld joint structured light positioning method based on image synthesis. The method comprises the following steps: S1, forward synthesizing a weld joint structured light image; and S2, weld joint morphology characteristics are solved reversely. According to the method, multi-type welding seam structured light image data are automatically generated by constructing a simulation environment, so that the adaptability and generalization ability of the model to different welding seam shapes can be improved; the weld seam image generated through simulation has known and accurate section geometric information, morphology measurement does not need to be conducted on a real workpiece, the quality and labeling precision of a training sample are improved, and then the overall weld seam positioning precision is improved; common noise interference in various welding scenes is introduced in the structured light image simulation process, so that the robustness and the anti-interference capability of an identification model in a complex environment are improved; the convolutional neural network is adopted to directly predict the weld joint section morphology, the traditional recognition process is simplified, error accumulation in the middle steps is avoided, and the weld joint positioning efficiency and precision are improved.